TWI248735B - Mobile ad-hoc network (MANET) including forward error correction (FEC), interleaving, and multi-route communication features and related methods - Google Patents

Abstract

A mobile ad-hoc network (MANET) (20) may include a source node (21), a destination node (22), and a plurality of intermediate nodes (23-28). The source node (21) may establish a plurality of routes (101-103) to the destination node (22) for transferring data therebetween, where each route passes through at least one of the intermediate nodes (23-28). The source node (21) may also encode a plurality of data packets (TLP1-TLPJ) using a forward error correction (FEC) encoding algorithm to generate error correction data for the data packets, interleave the data packets and error correction data, and distribute and send the interleaved data packets and error correction data across the routes (101-103) to the destination node (22). Furthermore, the destination node (22) may receive and deinterleave the interleaved data packets and error correction data. The destination node (22) may also decode the data packets based upon the error correction data using an FEC decoding algorithm to correct compromised data packets.

Description

1248735 坎, invention description: [Technical field to which the invention pertains] · The present invention relates to the field of mobile networks, in particular, / includes a pre-transmission error correction, and (I, left, wrong, and multipath communication characteristics) The purpose of the network and related methods. ^ The move [previous technology] During the exhibition period, the wireless network has developed more rapidly. Among them, the rapid development of the field - the mobile network (ΜΑΝΕΤ). In fact , ::Ik thinking network includes several geographical locations sharing a common radio channel. In the case of H and other other networks such as cellular or satellite network, the most popular features. That is, there is no fixed infrastructure. The network can be composed only of these mobile nodes, and when they are transmitted or received within the range of each other, they will cause a rapid 'on the fly' to build a network. The network does not depend on a particular node and can be dynamically adjusted as some nodes add A or a node out of the network. Due to these unique features, Routing protocols for managing data flows in MANET, which can accommodate changes in topology that occur from time to time. In recent years, two basic types of MANET routing protocols have been formed, namely reactive or "on-demand" agreements. And active or routing table-driven agreements. When a specific route to a target is required in response to a routing request, the reactive protocol collects routing information. Examples of reactive protocols include random request distance vector (AODV) routing, dynamics. Source Routing (DSR) and Temporary Sequential Routing Algorithm (TORA) O:\9l\9l208 DOC -6- 1248735 The type of protocol attempts to maintain all of the nodes in the network from each node in the network. 々— ^ ^ , “ie”, which is the consistent and up-to-date routing information. Such agreements are usually maintained in the same way as one of the routing information. π„, ή表, 亚和糟 transmits the updated data throughout the entire network in response to the network topology, and the construction of the network, in order to maintain one of the networks = view. Routes, such as the Target Sequence Distance Inward (DSDV) route, are disclosed in US Patent No. 5,412,654 to Perkins, 'Wireless Routing Protocol (WRp); and Cluster Head Gate Switching Routing ((10)R) The hybrid protocol of both the formula and the reactive method is a regional routing protocol (ZRp), which is not disclosed in the US Patent No. 6,304,556 of Haas. - The available data path between the source node and the shouting, but most of the financial rules are: a common feature system. In fact, the agreement usually selects one of the available paths from the source node and the target for communication. A single specified path. The path selection of the & path may be dependent on the path-dependent or multiple quality of service (QoS) factors, such as available bandwidth, delay, and the like. There are many difficulties in transmitting data in a dynamic MA bribe environment, and the reliability of the wireless communication link that constitutes the path between the road and the point is perhaps the most significant N〇va E in Cincinnati, Ohio. The NovaRoam Series 25〇 and 35〇 wireless routers, which are manufactured by the company, are a device that attempts to solve the link reliability problem, and implement pre-transmission error correction (fec) technology to improve the reliability of the communication link. As with most conventional communication systems and networks, these routers use FEC on the physical network layer. FEC only

O:\91\91208.DOC 1248735 Operates on a single hop, 浐之德,隹一绍, that is, at each hop, it receives the data transmission :: post-order decoding. Then, if one is needed, the information will be decoded again by the hopping point. U and hunting are received from the next point and then although such devices can provide - improved f-degrees in the event of a particular event 'but - these devices say, (d) SeduΓ provides sufficient error correction. For example, the information is lost. The present invention aims to provide a mobile random network (MANET) and related parties with enhanced: miscorrected features. Hunting is provided by a mobile network (MA bribe) to provide this and other objects, features and advantages in accordance with the present invention. The mobile network includes a source node and a plurality of intermediate nodes. The source node can establish a plurality of paths to the far target node. The paths are used for both: transmitting data, wherein each path passes through at least 2 nodes of the intermediate nodes. The source node may also encode a plurality of data packets by a forward error correction (FEC) encoding nasal algorithm to generate error correction data for the data packets and interleave the data packets and the error correction data, and The data packets and the error correction data are allocated and the data packets and the error correction data are sent to the target node via the paths. In addition, the target node may receive the interleaved packet and error correction data via a plurality of paths, and deinterlace the data packets and the like.

O\91\9l208.DOC 1248735 Error correction data.唁 俨 使用 Using FF ” can use the FEC decoding algorithm to decode the data packets based on the error correction data to correct the corrupted data packets. Therefore, the smuggling of the shackles and the distribution of the ## ms ^TFECm in the plural channels are again poor, even when the path is lost, 兮^anET is thus allowed to correct (10) damage Material data packets, and: When the new path is allowed, the transmission can be allowed to continue. The data is sealed by a plurality of paths to distribute the additional information related to the FEc code, thus reducing the increased bandwidth requirement. More specifically, the source Mingke point can interleave the data packets in columns so that each column contains the corresponding error correction data. For example, ° ’ 5 hai, the flat code FEC algorithm can be _ (10) ed Yang called the encoding algorithm, and the dream ^ ^ Romon code includes - corresponding to each column ... next: the data can therefore be coded word. In addition, the source P ′ q It divides the columns into a plurality of shelves, and sends the block to the target node via a plurality of paths, thereby encapsulating the data and the error correction data. ^ ^ The plurality of data packets may be transport layer packets and the like: ... the original point can be grouped into a plurality of network layer data packets by grouping the interleaved transport layer data seals In this way, (4) the error correction resource corresponding network layer packet bar two network layer data packet generation layer packet Lang points can be based on the network ground, the second J network layer data packet for deinterlacing. The 'far source' can also be inserted between these multiple packets before encoding.

〇\9 I\9I208.DOC I248735 7 packets are separated, and the target node can retrieve the data packets based on the packet delimiters after decoding. In addition, the source node may interleave the far data packet and the error correction data based on the at least one FEC parameter, and the source node may change at least one FEC parameter according to the service quality of the plurality of paths such as the beibei edge For example, the source node may interleave the data packets and the error %c by using at least one of block interleaving, convolutional interleaving, repetitive parental error, and pseudo-random interleaving. Moreover, by way of example, the source node can establish a plurality of paths using an active routing protocol, a reactive routing protocol, or a hybrid manet routing protocol. A method aspect of the present invention is for communicating between a source node and a "target node" in a manet, such as the method described above. The party includes a plurality of sets of the source node to the target. Path, used to pass between the two, where each path passes through the intermediate node/stomach point. In addition, the _ pre-transfer error correction (FEC) coding algorithm can be used at the source node. To encode a plurality of data packets, to generate error correction data by data packets, and to interleave the data packets and the error correction data in the case of ^ 丨 Μ Μ and at the point of the child. The method may also include assigning the data packet. Interleaving the data packet and the error correction data and transmitting the interleaved data packet and the error correction data to the target rib via the paths. 兮·g Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Μ Receiving the interleaved data packets and the error correction data, and the target node may also deinterlace the data packets and the error correction data. Included in the target node based on material error correction

O:\91\9I208.DOC -10- 1248735 The FEC decoding algorithm is used to decode the data packets to correct the corrupted data packets. The present invention will be described more fully hereinafter with reference to the accompanying drawings. However, the invention may be embodied in a variety of different forms and should not be construed as being limited to the specific embodiments set forth herein. Rather, the specific embodiments are provided so that this disclosure is thorough and complete, and such specific embodiments fully exemplify the scope of the invention. The same numbers in the entire drawings represent the same elements. The first month of the present invention is a mobile network (MANET) 20 comprising - source node 2 i, - target node 2 2 and a plurality of intermediate nodes 23-28 between them. The nodes 21_28 may be any type of mobile device capable of communicating within one, such as a computer, a personal digital assistant (PDA), etc., for example, including wireless communication devices 3, 31, and familiar with it. A device known to the skilled artisan. Of course, it should also be understood that in this application, if necessary, the nodes 2 - connectable to - the fixed communication infrastructure is wide, that is, the source node 21 and the target node 22 32, 33 may be used as appropriate. The controller cooperates with the wireless controllers of the respective controllers a 4 and 21 and the target node 22 to cooperate with the devices 30, 31 to transmit and receive the data point controller 32 illustratively including a six' source. _, and the target two..., the functions of the: solution;: two / described below. Lift

O\9l\9l208.DOC 1248735 For example, those skilled in the art will appreciate that the source node controller 32 and the target node controller 33 can be implemented using a microprocessor, memory, software, and the like. Further, the wireless communication devices 3Q, 31 can be wired, wireless local area network (LAN) devices, mobile telephone devices, and the like. It should also be understood that the intermediate nodes 23-28 preferably also include mating wireless communication devices/controllers, which are not shown in (d) for clarity of illustration: line communication devices/controllers. Of course, it should be understood that the portion of the path to the target node 22 also includes the wired infrastructure. The function performed by the node controller 32 establishes a plurality of paths between the source node 2i and the target node 22 for transmitting lean material therebetween. In the exemplary embodiment, three such routes are exemplarily shown, ie path ιοί (via nodes 24, 23 and 25), path 1 〇 2 (via nodes 24, 26 and 25) and path 103 (via 27) And 28). Of course, any number of paths can be used in accordance with the present invention. Each of the paths 1 〇 H 〇 3 illustratively includes a respective wireless communication link 29, and as indicated above, the material path passes through - or a plurality of nodes of the intermediate nodes 23·28. For example, the secret bribery path may include any number of intermediate nodes therein, depending on the proximity between nodes such as the network size. Each intermediate node along a path is typically referred to as a "hop", (h〇p), so sometimes the path through multiple intermediate nodes is referred to as a 'multiple hop, (mulit_h〇p Path. Note that although a relatively small number of intermediate nodes 21-28 are shown in the current example for clarity of description, any number of nodes may be included in the MANET 2® in accordance with the present invention.

O:\9I\91208.DOC -12· 1248735 Those skilled in the art should understand that the manner in which such paths l〇1_1〇3 are established will depend on the particular financial leg agreement implemented in the MANET 2〇. As indicated above, the path can be established using active protocols, reactive protocols, or its group σ. The proactive agreement continues to maintain the most recent routing option, while the reactive protocol explores the request path when the target node 22 needs to send the data. For example, any suitable MANET protocol that can be used to establish multiple paths, such as the μανετ protocol discussed above, can be used. Although these MANETs are still in their relatively early stages and have not yet adopted a uniform standard, it is likely that data communication within these MANETs will follow the Open Systems Interconnection (〇s) structure or variants. The 忒杈 type includes seven different hierarchical control layers, namely (from highest to lowest) application layer, presentation layer, session layer, transport layer, network layer, data link layer and physical layer. In the transmitting node, the control is passed from one layer to the next, starting from the application layer and continuing to the physical layer. The data is then sent through the path and when the data arrives at the target node, the data is processed in the reverse order of the hierarchy (i.e., from the physical layer to the application layer). In addition, the information corresponding to each particular layer is generally organized in the form of data packets. In accordance with the present invention, if the source node 21 has a data packet to send to the destination node, the FEC encoder 5 can optionally encode the data using a forward error correction (FEC) encoding algorithm. In general, FEC is used to process data before it is transferred using a predetermined algorithm that adds additional data or parity check bits to the data. Equivalence check

O:\9I\9I208.DOC -13- 1248735, the metasystem specifically does not count to correct any character block or t block error of the relevant data. Therefore, if the transmission was received incorrectly, the parity check bit can be used to check and repair the data. The source node 2 may be the originator of some or all of the data packets. That is, those skilled in the art should: Solution: The source node 21 can generate such data packets, pass a data from f, or can generate and deliver data. In addition, the fec code button 51 can encapsulate the data corresponding to any of the different layers of the 〇si model described above (or the same network layer corresponding to other suitable network models). Nevertheless, the invention is particularly suitable for encoding transport layer data packets. In particular, most wireless networks generally include a certain degree of cyclic redundancy check (CRC) in the data link layer/physical layer, so that (4) data transmission is wrong, however, at the lean link layer/physical layer. The CRC detects channel fading, bit, 曰$, etc., and the result is generally data packet loss and needs to be retransmitted at the data link layer, the transport layer, or the packet is permanently lost. However, by using - FEC encoding algorithm to encode the transport layer data: the packet 'can still use the data link layer/physical layer CRC to reduce channel errors ^ because the resulting higher layer of the target layer 22 can recover the resulting Any packet loss, and usually does not need to retransmit the data packet' is a very time consuming process. The encoding operation performed by the FEC pager 5 1 will be described in detail with reference to Fig. 2 . In the illustrated example, the source node controller preferably separates the packet from the symbol (beginning of the first packet), _ (before encoding the processing of the forwarding layer is not TLP1, TLP2, etc.) The second long and EOLP (the end of the last packet) and so on are inserted in the transport layer

O:\9I\91208.DOC -14- 1248735 Between the packets. The packet delimiter symbol is used to assist in reconstructing the transport layer data packet at the target node, as will be further described below. μ The FEC encoder 51 can encode the read (4) packet and the packet delimiter (if (10)) using various FEc encoding algorithms. A typical algorithm is the Reed_s〇i〇m〇n (Rs) coding algorithm. As is familiar to those skilled in the art, RS codes are not *effective in correcting erased symbols. - The RS code has a non-binary data symbol and (Ν_κ) non-binary parity check symbols (ie, N is the sum of the data symbols plus the = the length κ of the resulting parity check symbol) . Therefore, those skilled in the art should understand that the encoding rate is K / Ν and redundant or internal overhead (Ν-Κ) / Ν. In general, the RS symbol is selected as the m-bit symbol port (7) value, and in this case the Sigma code block length should satisfy the relation N < 2 m. In addition, the rs code minimum range is deducted N-K, which allows correction of any combination of N-K erasures in a single codeword. In other words, it is the so-called maximum range separable code, meaning that it can correct N-K erasures using the minimum internal addition of any possible code. In the present example, the transport layer data packets and their respective packet delimiters are preferably organized in column 35, each column being N bits long. Therefore, the bedding package and the data packet delimiter symbol will occupy κ bits before each column 35, and the last Κ-Κ bits of the column are unique to the particular one of the bits of the data in the column. The RS codeword is generated by the &3 coding algorithm to generate a unique RS codeword. Of course, those skilled in the art should be aware that these RS codewords can be attached anywhere in a column, or other

O:\9I\91208.DOC -15 · 1248735 The programme organizes such information so that it can be coded. In particular, when the FEC encoder 51 generates corresponding RS code words, the parent data error 50 can interleave the transport layer data packet and the packet delimiter into the columns 35, and in the columns The transport layer bevel packets and packet delimiters are again interleaved at the end. By performing interleaving across multiple RS codewords, thereby dispersing the effects of channel fading due to data congestion and accidental loss of links, channel fading and accidental loss of links will result in the use of familiarity in the door speed. Any suitable knowledge of the skilled artisan;: 3L error such as block parent error, which is illustrated in this current example. Other "3L class 5L erroneous techniques include convolutional interleaving, spiral interleaving, and pseudo-random interleaving. One is called "the same shame and the tiling is interleaved into the column 35, and then the information is distributed and passed through the paths. 1〇Μ〇3 sends the information to the target node 22. The way to distribute the information is particularly advantageous by dividing the columns into a plurality of blocks, so that each becomes a network layer packet (NLpi — ^荨荨), as shown in Figure 3, the mother-network layer packet therefore has the interleaved content of the packet. The transmission layer of the small U: This will lose several transport layer data if the network layer is lost during transmission. Packets and slags), which are all transport layer data packets in the _ column, Μ K - symbol block ^ every m day & ^,,, special temple: Of course, you can also use various bits/symbols Barriers to wide production. As will be described below in (4) one of 101-103, it is also said that the special path is lost in the mother-transport layer data packet.

O:\9I\9I208.DOC - 16- 1248735 - a bit of 7G, and the loss can be reconstructed at the source node based on the leftover of the 35th. - Corresponding column It should be understood that the column width is chosen to be an FFr. It seems that the bit is affected by the error correction essence of the tEC code ... s S code system - symbol error. , 杈 code. That is, its correction has a symbol quantum error, that is, whether it is consistent with a single bit 'θ 唬 唬 还是 还是 还是 还是 还是 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 If the coffee code is used as a binary error correction code (such as a binary convolutional meta or a binary BCH code), then a column width of one bit will preferably be selected. As with the transport layer packets, the source node controller 32 can generate a corresponding network layer packet header (not shown) for use by the target point n. For example, 'the network layer packet headers may include a first-class pointer (4)' a sub-stream ID (for example, for reserved path resources), a cross-block serial number, an interlace column number, - an agreement 1〇 (the higher level agreement used to identify the particular coffee application error), and/or the encoding/communication parameters used and/or other appropriate information, as known to those skilled in the art. 3 Those skilled in the art will also appreciate that if such paths 101-103' are established then the source node controller 32_ will typically set and reserve resources for each path. A method for facilitating the creation of multiple paths and transmitting data via multiple paths in a manet is described in U.S. Shen Jing Case No. h, 997, filed on August 8, 2002. This application has been assigned to the assignee and the entire contents of this application are hereby incorporated by reference. As described above, as part of the process of setting up and arranging the 伴 / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / The DOC -17- 1248735 number (i.e., M, N, K, m, etc.) is sent to the target node 22. In a : application, the source node controller 32 and the target node "will, negotiate" two & 乂: support FEC / parent error mode and parameters. It should be noted that the staggered/car j-plates used/sound delays must be considered as a path of delay: as part of the process - those familiar with the art should be aware of this. It should be understood from the above discussion that the mother and the interlaced parameters are related to each other. It is also true that the variables Nκτ, eight K, etc. will not only affect the encoding, but also control how the transport layer packets and the packet delimiters are interleaved into the columns μ. However, it is also necessary to change the source node controller 32 due to the different paths (8) (10) or the number of valid paths or other network qualities of the multiple products (Q〇s) metrics/measures. The FEC/interlace parameters. This can be implemented in different ways. For example, according to a certain MANET protocol, although a given path fails or the target node control state 33 can measure that the quality of service of a particular path is poor, and: when the source node controls n32, The source node 21 will receive a path failure notification. Thus, the source node controller 32 can change the new encoding/interlacing modes or parameters, or can negotiate the situation with the target node controller 33 via other signal exchange programs. It is also possible to use other appropriate methods known to those skilled in the art. For example, if a -specific path is lost, then an additional internal force may be required" and the wire therefore needs to inform the target node controller that the source node controller 32 has changed the encoding/interleaving parameters. Refers to the parent block's sequence number starting to indicate the new encoding parameter. That is, if there are 5 paths in use, the code rate should be 4/5 or less.

O:\9I\9I208.DOC -18- 1248735 and if /, there are 4 paths in use, then the code rate should be 3/4 or less, etc., as understood by those skilled in the art. Once the paths 1〇1_1〇3 have been established and the associated network layer packets and packet headers have been generated and the network layer packets and packet headers have been assigned by the respective paths, the source node controller 32 cooperates with the wireless communication device 30 to send the 4th network layer packet/packet header to the target via the corresponding path; When the source node 22 receives the network layer packets through the paths (8)_1〇3, the corresponding packet headers included therein indicate how the target node controller 33 sorts in the deinterleaver 52 to solve Interlace 0 It will be appreciated by those skilled in the art that when sufficient demultiplexed packets have been received in the deinterleaver 52, the reconstructed array of interleaved data can be deinterlaced (if necessary) and The FEC decoder 52 uses a fec decoding algorithm to perform the material, and the fresh EC decoding (four) method corresponds to the (four) algorithm used by the FEC encoding n51. (4) Ground, as described above - any uncorrected bit error, fading or link loss will result in discarding a number of other network packets, that is, the loss of the material to be paid for. The RS code word - the erasure symbol (or a bit of a two-bit read error correction code), once the Rs solution (four) charges the erasure payee ' then can be based on the packet delimiters Transport layer data packet. The requirement 'when implementing the MANET example, the lower cost but the more cost is consumed. Depending on the specific 20-year application of a particular MANET application, it is necessary to weigh or consider various design schemes. The RS encoding rate generally produces better performance.

O:\9I\9I208.DOC -19- 1248735 Internally attached. Similarly, the greater the length of the deinterlacer 5〇, the better the resulting efficiency, but the cost, additional delay. Regarding the decoding at the target node U. For those who are familiar with this technology, it should be understood that (4) horses (4) can be implemented in the form of hardware or software. Hardware implementations may be suitable for high-capacity rates because otherwise a large amount of computational load will be incurred at some point. In addition, there is still a trade-off between the size of the interleaver and the number of paths used. That is, having more paths allows the interleaver % size and the delay to be reduced. However, 'use a lot of roads in certain situations: it may not be practicable' because the paths may not exist or cannot be exhausted, so many paths are used so that each path carries one of the total flows." Dugong, 廿丨,, && & • 乂 乂 乂 邛 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , Coding rate 2 ---- 3 --~------- 0^667~ ~ 4 --~ one ~——-__ 5 -——________ _ 6 ~—------ - __〇·8333~____ η~, the mode of communication between the source node 21 and the target node 22 in the arguing. The method is initiated (block 40), and in block 41, a plurality of paths (8)·1〇3 from the source node 21 to the target node 22 are established. Furthermore, as described above, in block (4), the source node 2 i encodes a plurality of data packets using the -F E C encoding algorithm (as needed) to generate error correction data for the data packets, as well as at the source.

O:\91\9I208.DOC -20- 1248735 Node 2" also interleaves the data packets and error correction data (blocks are called. Thereafter, in block 44, the interleaved data packets and error corrections are allocated, And sending the material interleaved data packet and the error data to the target node 22 via the material path 1G1_H)3. In block 45, the interlaced data packet is received through the path 1G1. And error correction data, and then deinterlacing the data packets and the error correction data in block 46. As described above, in block 47, the FEC decoding algorithm is then used to transcode the data based on the error correction data. #数据封包(If #要), thereby correcting the damaged data packet, and then ending the method (block 48). The second method will be clear to the above, and the additional method aspect of the present invention will be clear, and therefore no longer This further description is familiar to those skilled in the art based on the foregoing description, and the present invention may be on or above the network layer as opposed to the method used by prior art devices such as the above-described n. Convenient to apply fec. Each __ node along the path (or path) to the target node 22 does not need to be decoded in the fourth (four) data. Instead, 'the system performs the solution at the target node, and the point 22 can be based on the edge. The solution is performed by the information received from the source node 21 to the plurality of paths of the target node. As in the conventional method, the present invention provides a bifurcation energy + ^ ^ ^

Eight Forces' is not allowed in the cross layer by using FEC. "Free, white technical experts should understand that by performing FEC and across multiple roads, the fourth method of falsification, the method of the invention is such that even if the road is completely lost One of them can also correct the corrupted data packets. Then, when reading & 7 creates a new path (if necessary), this is

O:\9I\9I208.DOC -21 . 1248735 The MANET 2G and method allow the transmission of the f f packet to be continued. In addition, although some additional internal additions are associated with the included fec parity check data, the internal addition advantageously spreads across multiple paths, thereby reducing the need to increase bandwidth. Furthermore, the above-described MANET and method of the present invention provides preliminary correction of such packets that are corrupted by link errors, fading, loss, etc., and the problems of the link errors, (4), loss, etc. are very significant in MANETs. The problem. The present invention provides a significant service. °Bei Jun point, especially for traffic with immediate delay demand, this traffic cannot allow retransmission of data in order to correct errors and losses. Those skilled in the art will also appreciate that the present invention can be implemented in numerous other network applications in addition to being implemented in MANET. For example, the data packets can be routed through the MANET and reach the final destination node via a fixed infrastructure, and vice versa. The present invention can also be advantageously utilized to handle wireless local area network (WLAN) handovers from a J access point in a relatively unobstructed and reliable manner to provide the desired quality of service. Similarly, the present invention is equally well suited for cellular handover between multi-base stations in a cellular network. The present invention can assist in reducing the telecommunications interruption experienced by the mobile phone user, and the electric=interruption will additionally occur when the mobile phone and the single-base station lose money. 0: In the future mobile communication architecture, the The present invention uses a different base station, and there is a problem, the link is to pass reliable data to - 曰μ non-<Beidingding" as known to those skilled in the art. The invention can also be implemented in a fixed wireless network with inaccessible paths, wherein

O:\9I\9I208.DOC -22 - 1248735 Links are susceptible to severe fading and loss. The present invention can significantly improve reliability and service quality. In particular, the FEC method outlined herein is particularly useful for improving the quality of service for instant traffic on the network, where the retransmission method is not allowed to utilize additional delays to provide packet reliability. Those skilled in the art will be aware of other applications based on the above discussion. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a mobile network (MANET) according to the present invention; FIG. 2 is a diagram for explaining interleaving of interleaver and data, packet delimiter, and error correction data. Figure 3 is a block diagram illustrating the allocation and transmission of interleaved data packets, packet delimiters, and error correction data from the interleaver of Figure 1. Figure 4 illustrates the use of a MANET in accordance with the present invention. Schematic diagram of the communication between one of the source nodes and one of the target nodes. [Character representation of the symbol] 20 Mobile network 21 Source node 22 Target node 23, 24, 25, 26, 27, 28 Intermediate node 30, 31 Wireless communication device 32, 33 controller 35 column 36 block 50 interleaver

O:\91\91208 DOC -23- 1248735 51 52 53 101, 102, 103 FEC Encoder Deinterleaver FEC Decoder Path O:\9I\91208.DOC -24-

Claims (1)

1248735 Picking up, applying for patent scope: L kinds of mobile random networks (MANET), including: a source point, a target node and a plurality of intermediate nodes; the δ Haiyuan node establishes a plurality of paths to the target point, the paths For transmitting data between the source node and the target node, each path passing through at least one intermediate node of the intermediate nodes; the original Lang also handles the plurality of data packets, and allocates the data packets and And the plurality of paths send the data packets to the target node; the target node receives the interleaved data packets via the plurality of paths and deinterleaves the data packets. 2. According to the mobile network of claim 1 of the scope of the patent application, wherein the source node also encodes the data packets using a forward error correction (FEC) coding algorithm, thereby generating error correction data. 3. An arbitrary network of actions in accordance with item 2 of the scope of the patent application, wherein the source node interleaves the data packets in a 2 interlace such that each column thus includes the = packet and the corresponding error correction data. 4. According to the action of the third paragraph of the scope of the patent application, the source node is allocated in this way by dividing the columns into a plurality of blocks and developing the columns to the target node via corresponding paths. And transmitting the interleaved data packets and the error correction data. 5. The mobile network according to claim 1 of the scope of the patent application, wherein the plurality of billet packets comprise transport layer data packets; and wherein the source node groups the interlaced transport layer data packets into a plurality of networks Layer seal O:\91\91208.DOC 1248735 This mode allocates the interleaved transport layer data packets. • A method of communicating between a source node and a point in a mobile network (MANET), the mobile network also includes a plurality of intermediate nodes, the method comprising: establishing a plurality of paths from the source node to the target point for transmitting data between the child point and the target node, each path passing through at least one intermediate node of the intermediate nodes; interleaving at the source node a plurality of data packets; allocating the interleaved data packets through the plurality of paths at a source or a point; ', τ < transmitting the allocated interleaved data packets from the source node to the target by a plurality of paths such as 4 And receiving, by the plurality of paths, the interleaved data packets at the target node, and deinterleaving the data packets at the target point. 7. The method of claim 6, further comprising: prior to interleaving the data packets, using a forward error correction _) coding algorithm to encode (4) the data packets to generate error correction data. 8. According to the method of claim 7 of the scope of the patent application, the interleaving comprises (iv) interleaving the processing of the data packets for each correction of the data. (10) thus including a corresponding error 9. According to the method of claim 8 of the scope of the patent application, the allocation includes dividing the column of interleaved data packets and error correction data into a plurality of blocks, and wherein transmitting comprises transmitting the blocks via the corresponding path 〃 4 target nodes. O:\9I\9I208.DOC -2- 1248735 1 〇 According to the method of claim 6, wherein the plurality of data packets comprise transport layer data packets; and wherein the allocation comprises interleaving the allocated The transport layer data packet is grouped into a plurality of network layer packets. O:\91\91208.DOC